The library is a collection place of various kinds of books. Arrangement of books by category called book shelving makes easier for customers to choose and find books. However, the location and arrangement of book categories becomes a problem in a library. Based on the book borrowing data, data mining was carried out to find out a book borrowed simultaneously by library visitor in one transaction. This can be solved by using the association rule technique and a priori algorithm. Possible combinations of borrowed books were based on certain rules and then tested whether the combination of items meets the minimum support requirements to create eligible rules. The results of this study were in the form of information about a combination of borrowed books for libraries to arrange the location of books according to categories that are often borrowed together.